Carcinogenesis in many mammalian systems can be dissected into at least two qualitatively different steps, i.e. initiation and promotion. Interference with tumor promotion is an important target for designing successful chemoprevention studies. In a variety of model systems, tumor promotion is directly correlated with polyamine biosynthesis, especially putrescine. A rate limiting step in mammalian polyamine biosynthesis is ornithine decarboxylase (ODC) induction, and inhibition of ODC induction may be important in developing rational approaches to use of chemopreventive agents in humans. Prostaglandin (PG) inhibitors suppress tumor formation by blocking PG synthesis and ODC induction. The use of combinations of chemopreventive agents may yield an enhanced effect as well as decrease the possibility of toxicities. We propose to test the hypothesis that measuring levels of PG biosynthesis and ODC induction in tissues after oral administration of a presumptive chemopreventive agent may serve as intermediate endpoints as well as biomarkers to determine biologically effective doses of the agents for use in large scale prevention trials in humans. A two-step study is proposed to determine a safe and tolerable dose of piroxicam (PXM) alone and in combination with difluoromethylomithine (DFMO) which will inhibit prostaglandin and polyamine biosynthesis, important steps in tumor promotion. In Step 1, PXM, a potent long- acting prostaglandin synthesis inhibitor, will be studied for its ability to inhibit the 12-O-tetradecanoylphorbol-13-acetate (TPA) induction of ornithine decarboxylase in 3 mm skin biopsies and thromboxane B2 (TXB2) release from platelets. The second step will study PXM alone and in combination with DFMO, another human chemopreventive agent, on the same biological endpoints. In addition during both Steps 1 and 2, urinary decarboxylated S-adenosyl-L- methionine (dc-SAM), which is proposed to be correlated with ODC inhibition by DFMO, will be measured.